73+ Built-in Tools for Agents

Capture screen with auto-scaling, grid overlay, and multi-monitor support. Vision-guided coordination system.

Get all connected monitors with resolution, position, and primary display status.

Click by grid cell reference (e.g., C3). Reliable clicking with lettered columns and numbered rows.

Click at precise coordinates. Supports left, right, middle buttons and double-click.

Move cursor to specific coordinates without clicking.

Click and drag from one point to another. Perfect for selections and drag-drop.

Scroll the mouse wheel at current or specified position.

Type text as if using a keyboard. Configurable keystroke delay for realistic input.

Press special keys (Enter, Tab, Escape, arrows, F-keys) with optional modifiers.

Execute keyboard shortcuts like Ctrl+C, Ctrl+V, Alt+Tab.

List all open windows with titles, positions, and sizes.

Bring a window to foreground by title or process name.

Read file contents with encoding support (utf8, base64, binary). Size limits for large files.

Write content to file. Creates file if it doesn't exist.

Apply patches or search-and-replace edits to existing files.

Search file contents using patterns. Fast grep-style searching.

Delete a file permanently.

Move or rename files.

Copy files to new locations.

Check if a file or directory exists.

List files and directories with metadata (size, modified date, permissions).

Create a new directory with recursive parent creation.

Execute shell commands with timeout and working directory. Detach mode for GUI apps.

Execute script files (bash, python, node) with custom interpreters and arguments.

Read text from system clipboard.

Write text to system clipboard.

Retrieve OS version, platform, architecture, and environment details.

List running processes with PID, name, CPU, and memory usage.

Terminate a process by PID.

Search the web and get structured results. Powered by search APIs.

Extract content from web pages. HTML to markdown conversion.

Make HTTP requests (GET, POST, etc.) with headers and body.

Create a new git worktree for parallel agent work. Automatic branch creation.

List all git worktrees in the repository.

Remove a git worktree when work is complete.

Switch active worktree for an agent.

Store information in vector memory. Facts, decisions, patterns, mistakes, successes.

Query vector memory with semantic search. Filter by type, importance, and topics.

Remove specific memories from vector store.

Get list of all available tools with descriptions and parameters.

Ask for help from other agents or the user. Posts to shared context.

Send a direct message to another agent in the session.

List all active agents in the session with their roles and status.

Request input or clarification from the user. Pauses execution until response.

Define a new multi-step workflow with agents and conditions.

Execute a saved workflow by ID.

List all saved workflows with metadata.

Check the execution status of a running workflow.

Define a background service with triggers (schedule, webhook, file watch).

List all registered services with status.

Manually trigger a service execution.

Enable or disable a service.

Remove a service permanently.

Save an agent configuration as a reusable template.

List all saved agent templates.

Retrieve full details of a specific template.